Calibrating the nonlinear matter power spectrum: Requirements for future weak lensing surveys

Uncertainties in predicting the nonlinear clustering of matter are among the most serious theoretical systematics facing the upcoming wide-field weak gravitational lensing surveys. We estimate the accuracy with which the matter power spectrum will need to be calibrated in order not to contribute appreciably to the error budget for future weak lensing surveys. We consider the random statistical errors and the systematic biases in P(k), as well as some estimates based on current N-body simulations. While the power spectrum on relevant scales (0.1 ≲ k/h Mpc−1 ≲ 10) is currently calibrated with N-body simulations to about 5-10%, in the future it will have to be calibrated to about 1-2% accuracy, depending on the specifications of the survey. Encouragingly, we find that even the worst-case error that mimics the effect of cosmological parameters needs to be calibrated to no better than about 0.5-1%. These goals require a suite of high resolution N-body simulations on a relatively fine grid in cosmological parameter space, and should be achievable in the near future.